KR20010032368A - Method of fine feature edge tuning with optically-halftoned mask - Google Patents

Abstract

A photolithographic mask is disclosed for optically transferring a lithographic pattern corresponding to an integrated circuit from a mask onto a semiconductor substrate using an optical exposure tool. The mask is a plurality of functional units corresponding to the elements forming the integrated circuit. And a plurality of invisible biasing segments disposed on at least one edge of the functional portion.

Description

METHOOD OF FINE FEATURE EDGE TUNING WITH OPTICALLY-HALFTONED MASK}

As semiconductor manufacturing technology is rapidly reaching the limits of optical lithography, to date, the state of the art processes have regularly produced ICs with functional features that exhibit critical dimensions (″ CD ″) below the exposure wavelength (″ λ ″). (The ″ threshold dimension ″ of the circuit is defined as the minimum width of the functional part or the minimum distance between the two functional parts.) For functional patterns designed to be smaller than λ, the optical proximity effect (OPE) is more severe, In fact it has been found to be unbearable during the leading edge sub-λ manufacturing process.

Optical proximity effect is a well known property of optical projection exposure tools. In particular, proximity effects occur when lithographically transferred circuit patterns that are fairly closely spaced are transferred to the resist layer on the wafer. As light waves of closely spaced circuit functions interact, they distort the final transferred pattern function. In other words, the diffraction causes adjacent functionalities to interact with each other in a way that produces pattern dependent variation. The size of the OPE on a given functional part depends on the placement of the functional part on the mask with respect to the other functional parts. The closer the functional parts are, the stronger the optical proximity effect between them.

One of the main problems caused by this proximity effect is unwanted variation in the functional CD. In leading edge semiconductor processes, this has a direct impact on wafer sort yield and the speed-binning of the final product, so strict control over the functional CD is the first goal of the manufacturing goal.

It is known that fluctuations in the CD of the circuit function generated by the OPE can be reduced by several methods. One such technique involves adjusting the illumination characteristics of an exposure tool. In particular, by carefully selecting the ratio of the numerical aperture (″ NAc ″) of the lighting capacitor to the numerical aperture (″ NAo ″) of the imaging objective (this ratio is referred to as the partial coherence ratio—σ). The degree of OPE can be operated to a predetermined range. The partial coherence ratio is defined as:

σ = (NAc) / (NAo)

In general, as σ increases, illumination coherence decreases. The less the illumination source interferes, the smaller the OPE is. In the extreme case where sigma is greater than 0.7, the OPE can be virtually minimized for feature pitches ("FP"), which range from separation to half separation, which is defined below. According to the rule of thumb, for modern exposure tools with NAo> 0.55, functional packing is often expressed in terms of the ″ pitch ″ for two adjacent functional parts. In addition to the description of the present invention as well as the following description, the functional part ″ pitch ″ is classified into the following four categories:

A) Dense function: FP <2λ,

B) semi-dense function: 2λ ≤ FP <3λ,

C) semi-separating function: 3λ ≤ FP <5λ,

D) Separation function: FP ≥ 5λ

Where FP = functional part CD + interval between functional part and functional part.

In addition to using relatively inconsistent illumination, as described above, the OPE can also be corrected by &quot; pre-calibration &quot; the mask function. This family of techniques is commonly known as optical proximity correction (OPC).

For example, U. S. Patent Application No. 5,242, 770 ('770 patent) assigned to the assignee of the present invention describes a method of using scattering bars (SBs) for OPC. The '770 patent describes a SB method that is quite effective at modifying the separation function to behave as if the function is a dense function. In this way, the depth of focus (DOF) for the isolation function is improved, thus significantly increasing the process range. Scattering bars (also referred to as intensity leveling bars or assist bars) are placed next to the separate functional edges on the mask (typically unresolved by the exposure tool) to adjust the edge intensity gradient of the separate edges. It has a calibration function. Preferably, the adjusted edge strength slope of the separation edge is matched with the edge strength slope of the dense feature edge, thus allowing the SB-supported separation function to have approximately the same width as the densely mounted function.

″ Scattering Bars ″ OPCs are an effective way to match separation and dense functions, while standard scattering bars (SSBs) carry more than the ″ optical weight ″ required when used in combination with the semi-separating function pitch. As a result, it is overcorrected if used in this context. In US patent application Ser. No. 08 / 808,587, filed Feb. 28, 1997, Applicants of the present invention deal with the need to make more precise CD corrections for semi-separated functions, ″ feature crowding ″. Describe the OPC method. The method described herein uses SBs with lighter optical weights so that the semi-separation function can be processed more accurately.

In particular, referring to FIG. 1, the ″ functional clustering ″ SB method of application 08 / 808,587 introduces two classes of new scattering bars: a thin SB 12 (″) having a width that is part of the width of the SSB 14. TSB ″), and halftone SB (″ HSB ″) 16, also known as dash-SB. HSB has the same width as SSB, but breaks into dashes using a halftone screen.

In the dense light between the functional parts 18, the light weight is directly proportional to the SB width. Therefore, by reducing the width of the TSB, the TSB can be clustered into a more compact functional space. However, the minimum width of the TSB is limited by the mask manufacturing process. In today's leading edge mask fabrication process, the minimum manufacturable TSB width is approximately 0.24 μm on the reticle (ie functional part). At the 1 × wafer scale, this is equivalent to a TSB width of 0.06 μm. As such, HSB should be used when an optical weight of 0.06 μm or less is needed.

HSBs have been developed to avoid the mask manufacturing limitations described above. Since the HSB can have the same width as the SSB, masks are easier to fabricate and inspect. Further, by adjusting the halftone period of the HSB (as described later with reference to FIG. 2), a desired optical weight can be obtained. For example, to obtain 50% optical weight, the halftone period must be 50% as shown in FIG.

Halftone Period (HTP) = d + s

while d = s,% H = (d / HTP) * 100% = 50%

Referring to Fig. 2, by adjusting the size ratio between d and s,% H can be changed to obtain the desired optical weight with respect to SSB. This HSB method extends the light weight below the minimum manufacturable width of the solid scattering bars added by the mask fabrication process. For example, using 0.1 μm SSB width as in reference, 25% HSB is equivalent to 0.025 μm width scattering bar. This is well below the current 0.06 μm TSB obtained by today's improved mask manufacturing process.

While the ″ functional clustering ″ OPC method of the 8 / 808,587 application is effective for CD control in semi-separated functional situations, it is not possible to physically insert an SB of the appropriate size between the functional parts in the semi-dense range. In the case of this type of functionality, OPE should be calibrated using ″ biasing ″.

A conventional functional biasing method requires changing the main functional CD by adding or subtracting a predetermined amount to or from the main functional part. For example, a +0.02 μm overall bias on a 0.18 μm main function portion changes the main function width to 0.20 μm. In the case of a semi-dense function, this type of bias correction is a sufficient OPC technique, as long as the mask writing tool can solve the required amount of bias.

In order to perform this +0.02 μm fine functional bias on the 4 × DUV wafer pattern, a bias of +0.08 μm (ie 0.02 μm × 4) is required on the reticle. Also, in order to preserve pattern symmetry and avoid positional movement of the functional part, half of this bias amount should be applied to each edge of each candidate functional part. The amount of bias required per side is then +0.01 μm. At 4 × this is only 0.04 μm. Imaging these patterns on a raster e-beam mask recording tool requires a 0.04 μm address unit that may require more than 20 hours of recording time for a standard 6 inch mask, despite being acceptable on modern e-beam machines. Shall. Clearly, this extended time period is either impractical or unacceptable from a production standpoint. A more acceptable production e-beam recording time is almost 3-6 hours per mask.

In addition, the use of an optical laser mask recording tool for such fine function biasing is impractical because it lacks sufficient resolution. Therefore, it is difficult to utilize the highly-fine biasing of the function in the half-dense function pitch range because of the high cost. Nevertheless, without some form of fine function biasing for the functions of the semi-dense functional pitch range, these pitch ranges cannot be tolerated in layout design rules.

As the semiconductor industry begins to use additional optical resolution enhancement techniques (such as alternating phase-shift masks), optical lithography is expected to produce minimum functional dimensions around 0.5λ. In this deep sub-λ circuit design rule, semi-dense functional parts (functional part pitch between 2λ and 3λ) are becoming more and more common. Thus, the semiconductor industry will soon require substantial OPC resolution for semi-dense features during the leading edge fabrication process. It is an object of the present invention to provide a solution to this problem.

TECHNICAL FIELD The present invention relates to photolithography, and in particular, to proximity correction features used in photolithography masks used for manufacturing semiconductor devices.

1 illustrates a ″ functional clustering ″ OPC method using thin scattering bars and halftone scattering bars described in US patent application Ser. No. 08 / 808,587.

FIG. 2 illustrates a method of defining a halftone period of a halftone scattering bar, as shown in FIG.

3 is an exploded view showing a portion of a mask formed using a typical embodiment of the halftone bias technique of the present invention providing positive bias;

4 is an exploded view showing a portion of a mask formed using a typical embodiment of the halftone bias technique of the present invention providing negative bias;

5 shows experimental results obtained utilizing the present invention to form a resist line.

Figure 6 illustrates the ability to adjust the micro-position of a given function using an asymmetric halftone bias technique in accordance with the present invention.

7 (a) and 7 (b) illustrate the ability to correct for lens aberration using the asymmetric halftone bias technique in accordance with the present invention.

8 illustrates an asymmetric halftone bias in accordance with the present invention applied on an altPSM parallel line pattern.

FIG. 9 shows intensity leveling obtained by applying the asymmetric bias shown in FIG. 8 to an altPSM parallel line pattern. FIG.

10 illustrates grid-snapping control usable using the halftone biasing technique of the present invention.

Accordingly, it is an object of the present invention to provide a cost effective and practical method of providing fine functional biasing for dense and semi-dense packetized functionalities to compensate for optical proximity effects.

In particular, the present invention provides a method for manufacturing a mask for optically transferring a lithographic pattern corresponding to an integrated circuit from a mask onto a semiconductor substrate using an optical exposure tool. The method includes forming a plurality of functional portions corresponding to elements forming an integrated circuit on a mask, and forming a plurality of non-resolvable biasing segments disposed on an edge of the at least one functional portion. It includes a step.

The present invention also relates to a lithographic mask for optically transferring a lithographic pattern corresponding to an integrated circuit from a mask to a semiconductor substrate by using an optical exposure tool. The mask includes a plurality of functional portions corresponding to the elements forming the integrated circuit, and a plurality of non-resolution biasing segments disposed on the edges of the at least one functional portion.

As will be described in detail below, the method and apparatus of the present invention provide significant advantages over the prior art. Most importantly, the present invention provides a halftone biasing method that results in significant cost savings by providing significantly fine functional size without the need to use unrealistically small e-beam address units or spot sizes. Indeed, the halftone biasing method allows nanometer scale functional biasing using current mask fabrication processes to allow OPC processing in situations of semi-dense and dense functionality where scattering bars are not available.

In addition, by asymmetrically applying the novel halftone biasing technique of the present invention to a given function, micro-position correction can be completed to compensate for lens-aberration-induced shift in resist function positioning.

The asymmetric halftone bias technique (″ AHB ″) also eliminates uneven atmospheric image intensities that can occur as a result of an unbalanced transmission between phase-shifting and non-phase-shifting in alternating phase shift masks (″ altPSMs ″). Effective for leveling

Another advantage of the AHB is that it can be utilized to modify and / or fix ″ grid-snapped ″ patterns that actually follow the originally intended placement.

Those skilled in the art will clearly recognize another advantage of the present invention from the following detailed description of exemplary embodiments of the present invention.

The present invention, together with the objects and advantages of the present invention, may be better understood with reference to the following detailed description and the accompanying drawings.

A novel biasing technique for controlling functional CD and functional pitch to provide OPC to dense and semi-dense packetized functional units as defined above will now be described in detail. A new biasing technique that utilizes optical halftones applied directly to the main functional edge provides significant fine sizing to the main functional portion.

3 is an exploded view of a photolithography mask, illustrating an exemplary embodiment to which the present invention is applied. As shown, each functional unit 30 includes a number of biasing segments 32 disposed on the edge 35 of a given functional unit 30. In the example of FIG. 3, the biasing segments 32 are arranged along two opposing edges 35 of each functional part 30. The biasing segments 32 forming a ″ tooth-like ″ pattern along a given functional portion 35 are shown as having a substantially square shape. However, in actual implementation, the biasing segment 32 exhibits a somewhat rounded configuration. The dimensions shown in FIG. 3 are defined as follows:

d = length of the given biasing segment

s = distance between adjacent biasing segments

w = the width of the given biasing segment

Typically, w is equivalent to the electron beam spot size used in the raster scan e-beam mask recorder. Spot sizes generally used vary from 0.05 μm to 0.125 μm for the state of the art 4 × DUV reticle.

As described in detail above, the functional pitch (″ FP ″), halftone period (″ HTP ″), and the percentage of halftones (″% H ″) of the biasing segment 32 are defined as follows:

FP = functional CD + functional-functional interval

HTP = d + s

% H = (d / HTP) * 100%

Importantly, in accordance with the present invention, the biasing segment 32 should not be disassembled by the optical exposure tool used in the photolithography process. This requires that the halftone periods of the biasing segments 32 disposed along the given edge 35 of the functional portion 30 maintain the sub-resolution. In order for the halftone period to maintain sub-resolution, the following equation must be satisfied according to Rayleigh's criterion:

HTP <k1 (λ / NAo)

Here, k1 = 0.61, lambda is equal to the wavelength of the optical exposure source, and NAo is equal to the numerical aperture of the imaging objective lens of the exposure source.

For example, for a 0.57 NAo KrF excimer laser DUV exposure tool with λ = 248 nm, the HTP should be less than 0.25 μm at 1 × wafer scale (ie 1.0 μm at 4 × reticle scale). In this example, when the halftone period HTP is stumbled to exceed 0.25 m, the flatness of the halftone is reduced, and a wave that stands out along the edge of the functional portion 30 starts to appear.

The biasing segment 32 may be applied as an external addition to the original function providing the positive bias, or as an internal cutout along the function edge providing the negative bias. 3 shows an example of both biases. 4 shows an example of negative bias. In particular, as shown in FIG. 4, a cutout 37 is formed along the functional edge, thereby forming a negative biasing segment.

3 means that the biasing segments 32 on the adjacent edges 35 of the functional part 30 (opposing edges of a given functional part or opposite edges of two adjacent functional parts) are not aligned with each other. An example of out-of-phase biasing segments is shown. As shown in FIG. 3, the biasing segments 32 of different phases are not aligned horizontally. Also note that the term ″ other phase ″ includes the environment when the biasing segment is shifted by half a period. Conversely, when the biasing segments 32 of FIG. 3 are in phase, the biasing segments 32 are aligned horizontally. However, as long as the above HTP rules are met, the same phase and other phase biasing segments 32 will provide equivalent CD results for the same% H.

Table 1 shows an example of how halftone bias can be precisely adjusted by adjusting% H (ie, the ratio between d and s). In the following example, the HTP rule is satisfied by fixing HTP to 0.25 μm, and the width of the halftone biasing segment is 0.020 μm per edge. Exposure conditions include 0.57 NAo KrF laser stepper with illumination σ = 0.80. The DUV resist utilized is Shipley's UV5 (0.61 μm thick) with Shipley's AR2 bottom anti-reflection coating on top of the polysilicon substrate.

number d s % H Resist CD bias One 0.05 μm 0.20㎛ 20% 0.189 μm + 0.009㎛ 2 0.10㎛ 0.15㎛ 40% 0.197 μm + 0.17㎛ 3 0.15㎛ 0.10㎛ 60% 0.205㎛ +0.025 μm 4 0.20㎛ 0.05 μm 80% 0.212 μm +0.032 μm 5 0.25 μm 0.00 μm 100% 0.220 μm + 0.040㎛

As indicated in the results shown in Table 1, the final CD of the function responds directly to the change in% H. At 100% H, this is equivalent to a total bias of 0.040 μm. For example, the DUV resist process is adjusted such that the measured resist CD bias is +0.040 μm. Thus, by applying the biasing segment of the present invention directly to a given edge of the functional part, the present invention can precisely adjust the CD of the functional part. In addition, as described in more detail below, a biasing segment may be applied to a single edge of the functional portion to ″ adjust ″ a single edge of the functional portion.

Although not all combinations are disclosed herein, as described above, the halftone biasing segment 32 may be applied in an additive or subtractive manner (positive halftone bias or negative halftone bias). An example of both biases is shown in Figure 3 and Table 1, where the biasing segment is utilized to increase the CD of the main function. In contrast, as shown in Fig. 4, the negative bias acts to reduce the CD of the main function. The negative biasing segment is formed in the same manner as the positive bias segment except that in the case of positive bias it is the opposite of adding the biasing segment to the edge of the functional part, and the biasing segment is removed by removing the section along the edge of the functional part. Is formed. In the case, the final edge shows the ″ this shape ″ appearance. In addition, the biasing segment is equally applicable to both dark and clear field masks.

FIG. 5 shows the experimental results that can be obtained when the illumination partial coherence σ is changed to affect the final functional CD obtained on the resist line. All three% H levels have the same tendency, i.e., the lower the sigma (i.e. the more interference) the higher the CD bias. See Table 1 for experimental conditions.

Partial coherence effects are useful for obtaining image intensities that are brighter than typical ″ holes ″ and ″ spaces ″. Higher hole and space CDs are expected to have smaller σ. As with the bright field function, the dark field function responds to the fluctuations at σ and% H as expected.

In addition, FIG. 5 appears for the given% H, the actual functional CD measured linearly tracks the partial coherence setting. Lower coherence settings (more coherent lighting), such as 0.3, will lead to higher CDs. CD, on the other hand, will be lower for higher partial coherence settings (more incoherent illumination). This property can be used to monitor the actual coherence settings used for wafer printing.

In addition, fine adjustment function position is made possible by applying different% H to the opposite edge of the function portion. For example, if the left edge of the functional part receives 20% H while the right edge of the functional part receives 80% H processing, the CD of the functional part will increase, but the center will shift slightly to the right, in fact the functional part ″ Micro-positioning ″. This effect is shown in FIG.

The ″ micro-positioning ″ adjustment effect can be utilized to prevent positional shifts due to lens aberrations. For example, in the case of a near-diffraction-limited exposure tool such as a state-of-the-art KrF excimer laser stepper, the amount of power half errors close to λ / 32 can be considered. The effect is calculated using the X-coma lens aberration of -0.03X (Z7 of Zernike polynominals). The same asymmetric halftone bias (″ AHB ″) correction factor as shown in FIG. 6 is used. The final resist pattern is shown in Figures 7 (a) and 7 (b).

In particular, Fig. 7 (a) shows the final position of the functional part caused by lens aberration without any correction. Figure 7 (b) shows a ″ calibrated ″ functionality that utilizes ″ micro-positioning ″ adjustments available as a result of the present invention. Indeed, the positioning correction effect is evident in the asymmetrically biased case shown in Figure 7 (b). In this example, lens aberration is corrected by shifting the function to the right using an asymmetric biasing segment.

In addition, the biasing segments of the present invention can be utilized to perform intensity leveling on alternating phase-shift masks (&quot; altPSM &quot;). In particular, FIG. 8 shows an altPSM line pattern having a number of main functional units 70. As shown, the functional part has a biasing segment 32 applied in an asymmetrical manner similar to the invention disclosed in conjunction with FIG. 7. The altPSM of FIG. 8 is designed for imaging of a 4 × KrF excimer laser exposure tool. . In this application, the function of the asymmetric biasing segment is to ″ level ″ the difference in intensity between the 0 degree region (0-space) and the π degree phase shift region (π-space). The intensity delta is known to be caused by transmission loss in π-space. Ideally, transmissions for 0 degrees and π space should be close to 100%. However, since a typical altPSM fabrication procedure uses an etching process to create a π space, the fabrication process often introduces a measurable amount of transfer loss in the etched area. Transmission losses in the π space as high as 10% were observed. This unbalanced transfer can create non-uniform sidewalls for resist function 70 formed between 0-space and π-space. The net effect is undesirable functional CD variation.

In order to correct this problem, a functional bias correction of approximately 40% H (0.02 μm) is needed (see FIG. 9). This is equivalent to a 0.008 μm edge bias. This bias degree, requiring a 0.032µm e-beam address unit, is not even possible with today's modern processes. On the other hand, using the halftone biasing technique of the present invention, this bias can be achieved using 0.1 μm e-beams at 4 × reticles.

In the application of the asymmetric biasing segment applied to altPSM as shown in FIG. 8, a 40% halftone bias is applied, w = 0.02 μm, d = 0.10 μm and HTP = 0.25 μm. The biasing segment only applies to functional edges close to zero-space. There is no halftone bias in the π-space. The main functional part CD is 0.14 mu m, and the main functional part pitch (&quot; FP &quot;) is 0.30 mu m.

Referring to FIG. 9, the application of an asymmetric biasing segment (“AHB ″) to altPSM works to eliminate intensity fluctuations between 0-space and π-space. Note that the above-described example of intensity leveling by AHB is applicable to the dense function (ie, FP <2λ, where λ for KrF excimer laser is 0.248 μm). Is available at. In the case of semi-separation and separation functions, the strength leveling requirements can be satisfied with scatter bars.

The AHB technique of the present invention can be utilized to shift the ″ grid-snapped ″ functionality to the original intended placement position. 10 shows a comparison of uncorrected (″ grid-snapped ″) resist patterns to AHB biased resist pattern contours. Grid-snapping occurs when the last e-beam address unit is not the full product of the original drawing grid. Under this situation, the functional edge will snap to the nearest multiple grid. This ″ grid-snapping ″ may shift the exact placement of the function from the intended position. Obviously, this is not desirable when manufacturing any leading edge IC. Using the AHB technique of the present invention as described above, it is possible to change the snapped function placement so that the original placement can be retrieved or closer to the need without reclassifying to a smaller e-beam address size. In particular, an asymmetric biasing segment with% H necessary to bias the edge back to the intended position is applied to the edge of the ″ snapped ″ function.

As mentioned above, the methods and apparatus of the present invention provide significant advantages over the prior art. Of particular importance, the present invention provides a halftone biasing method that provides significantly fine functional sizing without the need to use unrealistically small e-beam address units or spot sizes, resulting in significant cost savings. Indeed, the halftone biasing method of the present invention allows for OPC treatment in semi-dense and dense functional situations where scattering bars cannot be utilized by allowing nanometer-scale functional biasing utilizing current mask fabrication processes.

In addition, by applying the novel halftone biasing technique of the present invention asymmetrically to a given function, micro-positional correction can be achieved to correct lens-aberration-induced shifts in resist function positioning.

Asymmetric halftone bias technique (AHB) also levels non-uniform atmospheric image intensities that can occur as a result of unbalanced transmission between phase-shifting and non-phase-shifting in alternating phase-shift masks (altPSM). Effective at

Another advantage of AHB is that it can be utilized to change and / or correct a `` grid-snapped '' pattern to virtually follow the originally intended placement.

The partial coherence rate σ interacts with the performance of the halftone biased function. Thus, with appropriate adjustment, the combination of these effects can be used to modulate the light and dark field functional CDs, or the functional CD can be used to monitor partial coherence settings.

While certain specific embodiments of the invention have been disclosed, it is noted that they may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. Accordingly, the embodiments of the invention are to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than the foregoing description, and meaning and scope equivalent to the scope of the claims. Includes all changes within

Claims (17)

A method of manufacturing a mask for optically transferring a lithographic pattern corresponding to an integrated circuit from a mask onto a semiconductor substrate using an optical exposure tool, the method comprising: Forming a plurality of features formed on the mask and corresponding to the elements forming the integrated circuit, and Forming a plurality of non-resolvable biasing segments disposed on an edge of at least one of the functional units. Mask manufacturing method comprising a. The method of claim 1, And wherein said plurality of invisible biasing segments disposed on at least one edge of said functional portion are substantially equally spaced from each other by a distance-s. The method of claim 2, Each of the plurality of resolvable biasing segments has a predetermined width-w and a predetermined length-d, wherein the plurality of resolvable biasing segments have a critical halftone period equal to d + s. The mask manufacturing method characterized by the above-mentioned. The method of claim 3, And the mask is utilized in a photolithography process comprising an exposure source having an associated wavelength and an imaging objective having an associated numerical aperture NAo. The method of claim 4, wherein And wherein said critical halftone period of said non-degradable biasing segment is less than (λ / NAo) multiplied by 0.61. The method of claim 5, And wherein said plurality of functional portions exhibits a functional pitch of less than 3λ. The method of claim 5, And a plurality of said non-resolvable biasing segments are formed on at least two edges of at least one of said plurality of functional parts. The method of claim 5, And a plurality of said non-degradable biasing segments are formed on at least two edges of said plurality of said functional portions. A photolithography mask for optically transferring a lithographic pattern corresponding to an integrated circuit from a mask onto a semiconductor substrate using an optical exposure tool, A plurality of functional units corresponding to the elements forming the integrated circuit, and A plurality of invisible biasing segments disposed on an edge of at least one of the functional units A photolithography mask comprising a. The method of claim 9, And the plurality of invisible biasing segments disposed on at least one edge of the functional portion are substantially equally spaced from each other by a distance-s. The method of claim 10, Wherein each of said plurality of invisible biasing segments has a predetermined width-w and a predetermined length-d, said plurality of invisible biasing segments defining a critical halftone period equal to d + s. Photolithography mask. The method of claim 11, And the mask is utilized in a photolithography process comprising an exposure source having an associated wavelength and an imaging objective having an associated numerical aperture NAo. The method of claim 12, And wherein said critical halftone period of said non-degradable biasing segment is less than (λ / NAo) multiplied by 0.61. The method of claim 13, And wherein said plurality of functional portions exhibits a functional pitch of less than 3λ. The method of claim 13, And a plurality of said non-resolvable biasing segments are formed on at least two edges of at least one of said plurality of functional parts. The method of claim 13, And a plurality of said non-resolvable biasing segments are formed on at least two edges of said plurality of said functional portions. The method of claim 13, And the bias amount seen by the predetermined edge of the functional portion is changeable by adjusting the ratio between d and s seen by the plurality of incomparable bias segments disposed on the predetermined edge.